Monday, December 28, 2009

The Permanent Portfolio

In the last entry, we discussed the fiscal discipline imparted by the four-part "Crashproof" structure of a limited-purpose checking account, untouchable savings account, discretionary slush fund, and investment clearinghouse account. In this post, we'll specifically look at the investment account and discuss a concept that the late author Harry Browne termed the "Permanent Portfolio".

Harry Browne was a very intriguing character. A free market/Austrian economics disciple and former Libertarian Party presidential candidate, Browne taught himself economics and finance and worked as an investment advisor, specializing in the Holy Trinity of alternative investment strategies (AIS), asset protection structures, and offshore banking, for about three decades. One of his books, How I Found Freedom in An Unfree World, is probably the coolest self-help guide I have ever read.

(Harry Browne)

The essence of Browne's Permanent Portfolio concept is to create a conservative, all-weather/all-terrain investment system that allows the owner to sleep at night, regardless of what is happening in the world. While part of an investment clearinghouse account can be deployed towards active management styles and strategies that attempt to time the market to some extent (which of course is what a hedge fund manager does), the Permanent Portfolio would represent the "passive" component of the overall program. He wanted it to combine three features: safety ("it should protect you against every possible economic future. You should profit during times of normal prosperity, but you should also be safe ((perhaps even profit)) during bad times"); stability ("even in the worst possible circumstances, the portfolio's value should drop no more than slightly---so that you won't panic and abandon it"); and simplicity (the portfolio should be easy to maintain).

Browne breaks the macroeconomic landscape down into four general scenarios: prosperity (bull markets), inflation (rising prices, probably due to loose monetary policies previously discussed on this blog), tight-money (Federal Reserve response to inflation), and deflation (prices decline, purchasing power of cash grows, can be associated with a depression if it occurs rapidly due to demand shocks). He makes no attempt at predicting which scenario is more likely from a probabilistic standpoint; he considers them equally probable for planning purposes.

Harry Browne: "To be protected in all circumstances, each economic environment must have at least one investment in the Permanent Portfolio that responds well to it."

Browne then splits his Permanent Portfolio into four asset classes: stocks, bonds, gold, and cash. Stocks would do very well during a prosperity scenario, bonds do pretty well during prosperity (as interest rates fall) and deflation (as interest rates are aggressively cut down by a desperate central bank), gold does very well in an inflationary environment (tends to do fairly poorly in the other three), and cash earns a premium during a tight money situation and in a deflation (cash is neutral during prosperity and a loser during inflation). I should note that Browne's scenario narratives generally do not involve an absolute, final economic catastrophe, like an actual sovereign default by the United States. This is probably reasonable, since the US government would resort to debt monetization in that situation, and try to inflate its way out of its debts through the printing press (some believe this exact scenario has become likely, and we'll devote a blog entry specifically to discussing it).

Browne's asset allocation system is to divide investments equally across the four classes: each gets 25%. This form of allocation is called "1/N" in investment literature because an allocation receives a funding level of 1 divided by the number of investment options (in this case 4). In contrast to 1/N, much of Modern Portfolio Theory rests on the notion that statistical information regarding asset performance, specifically expected return, standard deviation (used as a proxy for risk), and correlation can be fed into an optimization engine and an ideal portfolio mix for some target level of "risk" (i.e., standard deviation)---say, 59.6% allocated to stocks and 40.4% allocated to bonds---can be constructed and implemented.

Snobs will say that 1/N is remedial, almost vulgar in its simplicity. In real-world testing, however, 1/N has routinely outperformed the more quantitatively sophisticated mean-variance portfolio optimization, and this is true for two reasons: 1) financial market prices do not follow the Gaussian statistical distribution that would give terms like standard deviation meaning, and correlations are unstable and tend to increase when the economy is stressed; 2)in trading system design, optimization of this naive, fragile kind is called "overfitting to the curve" and is considered very foolish. In fact, it appears that even the founder of the mean-variance portfolio optimization revolution, Harry Markowitz, ultimately used 1/N when it came to making real decisions about his own investment portfolios, and the Black Swan himself, Nassim Taleb, likes it.

In terms of periodic rebalancing, Brown recommends looking at the net asset value of the portfolio and modifying it if any particular asset rises to over 35% of the total NAV, or if any has fallen to below 15%. An investor basically makes sure that each investment has an equal allocation. In his words, "To rebalance the portfolio to its original percentages, just sell enough of the leading investments to reduce each to 25% of the total value. Use the proceeds from those investments to buy more of the investments that have fallen below 25%."

The intuition might be that Permanent Portfolio would never make any money, since gains in one asset class would be offset by losses in another. In practice, the gains in the winners reliably tend to be larger than the losses in the losers, hence a positive return is achieved over time.

There are various ways to go about constructing a Permanent Portfolio, but the key elements in my humble opinion are: A) exposure to the US S&P 500 or Russell 3000 index with as little friction (fees, transaction costs) as possible, which means purchasing an index ETF; B) exposure to the US Treasury 10-year; C) deciding whether to invest in a gold fund, to physically hold gold bullion, or both (pros and cons of various methods of gold investment will be left for a future discussion, but for now I will say that buying gold mining company stocks is not the same as having true exposure to the commodity itself); and D) deciding whether the cash allocation should be in addition to the cash already held in a contingencies/savings account (the cash will probably be put in a Treasury Money Market Fund or Money Market Fund mixed with three-month T-bills).

Alternatively, there is a mutual fund family based on the Permanent Portolio (but with other allocations within it). One member, PRPFX, has achieved a 5-year return of 8.62%. If you followed Browne's recommended portfolio mix to the letter since 1972 (after the US left the gold standard so that the printing press could be more fully put to work), you would have achieved a compound annual growth rate of about 9.7%, which defeats a (much riskier) 100% US stock market allocation over the same period. The thing actually achieved a positive return (+2%) last year; I think the worst year was 1981, when it lost 4%.

For further reading: I have enjoyed all of Browne's books, particularly the aforementioned, philosophical "How I Found Freedom in An Unfree World", and managed to get a couple of them signed by the author before he died. The most concise discussion of the Permanent Portfolio is found in the quick, accessible "Fail-Safe Investing", although a Google search under "Browne Permanent Portfolio" will yield all kinds of interesting tidbits.


  1. I'd be curious to know how you treat funds locked in 401k accounts and Roth IRAs.

  2. You'd probably have to go with a self-directed IRA option for the gold in particular (I believe gold bullion was approved for IRAs in the late 90s, but would have to check). We have some clients who place IRA money with FISERV/Lincoln Trust custodians in order open managed futures accounts with us.

    The Permanent Portfolio mutual fund would obviously be an easy way to get this kind of exposure.

  3. Nearly two years later, any new thoughts on PRPFX? I've been reasonably pleased with return since I entered last year.

  4. It would be sad to have your business greatly affected by the divorce issue.
    nevada llc